From: Sabrina Dubroca <sd@queasysnail.net>
To: Paolo Abeni <pabeni@redhat.com>
Cc: Marek Mietus <mmietus97@yahoo.com>,
netdev@vger.kernel.org, kuba@kernel.org, Jason@zx2c4.com
Subject: Re: [PATCH net-next v5 00/11] net: tunnel: introduce noref xmit flows for tunnels
Date: Tue, 17 Mar 2026 12:28:10 +0100 [thread overview]
Message-ID: <abk6yhfnry2NzkZh@krikkit> (raw)
In-Reply-To: <89df6470-027b-4d86-98a1-72042ca819ed@redhat.com>
2026-03-17, 12:02:30 +0100, Paolo Abeni wrote:
> On 1/13/26 5:29 PM, Marek Mietus wrote:
> > Currently, tunnel xmit flows always take a reference on the dst_entry
> > for each xmitted packet. These atomic operations are redundant in some
> > flows.
> >
> > This patchset introduces the infrastructure required for converting
> > the tunnel xmit flows to noref, and converts them where possible.
> >
> > These changes improve tunnel performance, since less atomic operations
> > are used.
> >
> > There are already noref optimizations in both ipv4 and ip6.
> > (See __ip_queue_xmit, inet6_csk_xmit)
> > This patchset implements similar optimizations in ip and udp tunnels.
> >
> > Benchmarks:
> > I used a vxlan tunnel over a pair of veth peers and measured the average
> > throughput over multiple samples.
> >
> > I ran 100 samples on a clean build, and another 100 on a patched
> > build. Each sample ran for 120 seconds. These were my results:
> >
> > clean: 71.95 mb/sec, stddev = 1.71
> > patched: 74.92 mb/sec, stddev = 1.35
> >
> > TL;DR - This patchset results in a 4% improvement in throughput for
> > vxlan. It's safe to assume that we might see similar results when testing
> > other tunnels.
>
> IMHO this performance delta is not enough to justify this amount of
> changes.
>
> Additionally, the measured impact of removing the dst_hold/dst_release
> does not fit with my direct experience on the same matter: it should be
> below noise level in practice, as dst are per-cpu and and no
> contention/false sharing is expected in a good setup.
>
> I think you are observing larger impact because in the veth test
> dst_release can happen on a remote CPU. Note that this setup (vxlan over
> veth) is not very relevant in practice.
>
> I'm sorry I'm not applying this series.
This is probably fine for everybody who's been following this series,
but you answered v5 (from January) instead of v8 (posted last week).
--
Sabrina
prev parent reply other threads:[~2026-03-17 11:28 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20260113162954.5948-1-mmietus97.ref@yahoo.com>
2026-01-13 16:29 ` [PATCH net-next v5 00/11] net: tunnel: introduce noref xmit flows for tunnels Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 01/11] net: dst_cache: add noref versions for dst_cache Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 02/11] net: tunnel: convert iptunnel_xmit to noref Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 03/11] net: tunnel: convert udp_tunnel{6,}_xmit_skb " Marek Mietus
2026-01-18 0:36 ` [net-next,v5,03/11] " Jakub Kicinski
2026-01-20 13:25 ` Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 04/11] net: tunnel: allow noref dsts in udp_tunnel{,6}_dst_lookup Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 05/11] net: ovpn: convert ovpn_udp{4,6}_output to use a noref dst Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 06/11] wireguard: socket: convert send{4,6} to use a noref dst when possible Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 07/11] net: tunnel: convert ip_md_tunnel_xmit " Marek Mietus
2026-01-13 16:29 ` [PATCH net-next v5 08/11] net: tunnel: convert ip_tunnel_xmit " Marek Mietus
2026-01-13 16:36 ` [PATCH net-next v5 09/11] net: sctp: convert sctp_v{4,6}_xmit " Marek Mietus
2026-01-18 0:36 ` [net-next,v5,09/11] " Jakub Kicinski
2026-01-13 16:36 ` [PATCH net-next v5 10/11] net: sit: convert ipip6_tunnel_xmit to use a noref dst Marek Mietus
2026-01-18 0:36 ` [net-next,v5,10/11] " Jakub Kicinski
2026-01-13 16:36 ` [PATCH net-next v5 11/11] net: tipc: convert tipc_udp_xmit " Marek Mietus
2026-03-17 11:02 ` [PATCH net-next v5 00/11] net: tunnel: introduce noref xmit flows for tunnels Paolo Abeni
2026-03-17 11:28 ` Sabrina Dubroca [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=abk6yhfnry2NzkZh@krikkit \
--to=sd@queasysnail.net \
--cc=Jason@zx2c4.com \
--cc=kuba@kernel.org \
--cc=mmietus97@yahoo.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox